Real-Time Detecton of Periodic Motion Systems and Methods

1. A system comprising: an imaging module configured to capture video of a scene; and a logic device configured to communicate with the imaging module, wherein the logic device is configured to: receive a video sequence of the scene from the imaging module, wherein the received video sequence comprises a sequence of frames and comprises one or more video regions that are present in each frame and are pixel-wise consistent between successive frames of the received video sequence; determine at least one periodicity deviation corresponding to at least one of the one or more video regions based, at least in part, on a periodicity deviation of an image at each of a plurality of pixels in the at least one video region, wherein the periodicity deviation corresponding to at least one of the one or more video regions measures a deviation of the video sequence from being T-periodic for a preselected period T; and issue an alert based, at least in part, on the determined at least one periodicity deviation.

2. The system of claim 1 , wherein the logic device is configured to: determine a spatial average of the received video sequence; and de-trend the received video sequence by subtracting the determined spatial average from each frame of the received video sequence, prior to the determining the periodicity deviation; wherein the periodicity deviation of an image at each of a plurality of pixels is based, at least in part, on a change of the image at each of the plurality of pixels between a first time and a corresponding second time preceding the first time by the preselected period T as the first time varies at least within an interval equal to the preselected period T.

3. The system of claim 1 , wherein the logic device is configured to: apply a low pass filter to the received video sequence, prior to the determining the periodicity deviation.

4. The system of claim 3 , wherein: the low pass filter comprises a linear low pass filter; and the linear low pass filter is implemented by a separable two dimensional Bernoulli filter.

5. The system of claim 1 , wherein: the one or more video regions comprise a plurality of rectangular video regions; each rectangular video region at least partially overlaps at least one other rectangular video region; and each rectangular video region comprises a height or width between 5 pixels and one half a corresponding full height or width of a frame of the received video sequence.

6. The system of claim 1 , wherein: the one or more video regions comprise a plurality of overlapping rectangular video regions; and the plurality of overlapping rectangular video regions comprises a corresponding plurality of positions and/or sizes configured to completely cover each frame of the received video sequence.

7. The system of claim 1 , wherein: the determining the periodicity deviation comprises determining a discretized, grey level translation invariant, normalized periodicity deviation corresponding to the at least one video region based, at least in part, on the at least one video region and the preselected period T; and the temporal length of the received video sequence is at least twice the preselected period T.

8. The system of claim 1 , wherein the one or more video regions comprise a plurality of video regions, the at least one periodicity deviation corresponding to the at least one of the one or more video regions comprises a plurality of periodicity deviations corresponding respectively to the plurality of video regions, and the logic device is configured to: determine the plurality of periodicity deviations corresponding respectively to the plurality of video regions based, at least in part, on the corresponding respective video regions and one or more preselected periods including the preselected period T or a preselected range of periods including the preselected period T; and issue the alert based, at least in part, on the determined plurality of periodicity deviations, the one or more preselected periods or the preselected range of periods, and a periodicity deviation threshold.

9. A method comprising: receiving a video sequence of a scene from an imaging module, wherein the received video sequence comprises a sequence of frames and comprises one or more video regions that are present in each frame and are pixel-wise consistent between successive frames of the received video sequence; determining a periodicity deviation with respect to at least one of the one or more video regions based, at least in part, on a periodicity deviation of an image at each of a plurality of pixels in the at least one video region, wherein the periodicity deviation corresponding to at least one of the one or more video regions measures a deviation of the video sequence from being T-periodic for a preselected period T; and issuing an alert based, at least in part, on the determined periodicity deviation.

10. The method of claim 9 , further comprising: determining a spatial average of the received video sequence; and de-trending the received video sequence by subtracting the determined spatial average from each frame of the received video sequence, prior to the determining the periodicity deviation; wherein the periodicity deviation of an image at each of a plurality of pixels is based, at least in part, on a change of the image at each of the plurality of pixels between a first time and a corresponding second time preceding the first time by the preselected period T as the first time varies at least within an interval equal to the preselected period T.

11. The method of claim 9 , further comprising: applying a low pass filter to the received video sequence, prior to the determining the periodicity deviation.

12. The method of claim 11 , wherein: the low pass filter comprises a linear low pass filter; and the linear low pass filter is implemented by a separable two dimensional Bernoulli filter.

13. The method of claim 9 , wherein: the one or more video regions comprise a plurality of rectangular video regions; each rectangular video region at least partially overlaps at least one other rectangular video region; and each rectangular video region comprises a height or width between 5 pixels and one half a corresponding full height or width of a frame of the received video sequence.

14. The method of claim 9 , wherein: the one or more video regions comprise a plurality of rectangular video regions; and the plurality of rectangular video regions comprises a corresponding plurality of positions and/or sizes configured to completely cover each frame of the received video sequence.

15. The method of claim 9 , wherein: the determining the periodicity deviation comprises determining a discretized, translation invariant, normalized periodicity deviation corresponding to the at least one video region based, at least in part, on the at least one video region and the preselected period T; and the temporal length of the received video sequence is at least twice the preselected period T.

16. The method of claim 9 , wherein the one or more video regions comprise a plurality of video regions and the at least one periodicity deviation corresponding to the at least one of the one or more video regions comprises a plurality of periodicity deviations corresponding respectively to the plurality of video regions, the method further comprising: determining the plurality of periodicity deviations corresponding respectively to the plurality of video regions based, at least in part, on the corresponding respective video regions and one or more preselected periods including the preselected period T or a preselected range of periods including the preselected period T; and issuing the alert based, at least in part, on the determined plurality of periodicity deviations, the one or more preselected periods or the preselected range of periods, and a periodicity deviation threshold.

17. A method comprising: receiving a video sequence of a scene from an imaging module, wherein the received video sequence comprises one or more video regions that are pixel-wise consistent between successive frames of the received video sequence; determining a periodicity deviation with respect to at least one of the one or more video regions based, at least in part, on the at least one video region; and issuing an alert based, at least in part, on the determined periodicity deviation; wherein the at least one periodicity deviation comprises first and second periodicity deviations corresponding to respective first and second identified periods associated with a biological periodic motion, the method further comprising: determining the first periodicity deviation corresponding to the first identified period is greater than or equal to a periodicity deviation threshold corresponding to the biological periodic motion; determining a first probability associated with the first identified period and the biological periodic motion is greater than a probability threshold associated with the biological periodic motion; determining the second periodicity deviation corresponding to the second identified period is less than the periodicity deviation threshold corresponding to the biological periodic motion; and issuing the alert based, at least in part, on the second periodicity deviation, to indicate a presence of the biological periodic motion in the video sequence.

18. The method of claim 17 , wherein: the method comprises updating a probability table associated with a set of identified periods and the biological periodic motion; the set of identified periods comprises at least the first and second identified periods; the probability table comprises at least the first probability associated with the first identified period and the biological periodic motion and a second probability associated with the second identified period and the biological periodic motion; and the updating the probability table comprises increasing the second probability and decreasing at least the first probability in the probability table.

19. A system comprising: an imaging module configured to capture video of a scene; and a logic device configured to communicate with the imaging module, wherein the logic device is configured to perform the method of claim 17 .

20. The system of claim 19 , wherein: the logic device is configured to update a probability table associated with a set of identified periods and the biological periodic motion; the set of identified periods comprises the first and second identified periods; the probability table comprises at least the first probability associated with the first identified period and the biological periodic motion and a second probability associated with the second identified period and the biological periodic motion; and the updating the probability table comprises increasing the second probability and decreasing at least the first probability in the probability table.